Introduction

Agromyces aureus AR33T is a type strain belonging to the Microbacteriaceae family, Actinobacteria phylum [1]. It is a heavy metal resistant bacteria that was isolated from the rhizosphere of a willow tree ( Salix caprea L.) grown in a heavy metal contaminated site (Arnoldstein, Austria). Among other bacteria isolated from the same source, AR33T was able to significantly increase the extractability of zinc and cadmium from a contaminated soil [2]. Moreover, the inoculation of AR33T in combination with the fungus Cadophora finlandica caused an increase of zinc and cadmium concentration in the shoots of Salix caprea L. plants growing in a heavy metal contaminated soil [3]. Based on these interesting features and the fact that the Agromyces genus is still a relatively unexplored genus, we decided to sequence the whole genome of A. aureus AR33T to gain insights in this genus and the heavy metal resistance and immobilization and mobilization mechanisms. At the time of writing (June 2016), 27 species of the Agromyces genus have been recognized and only nine draft genomes are available in the NCBI database. Here, we present the first complete genome sequence of an Agromyces species, A. aureus AR33T and a comparative analysis with other Agromyces spp. and related members of the class Actinobacteria .

Organism information

Classification and features

A. aureus AR33T is a Gram-positive bacterium having yellow-pigmented colonies (Fig. 1a). Cells are rod shaped and can form curved hyphae (Fig. 1b). Phylogenetic analysis based on 16S rRNA genes of other Agromyces strains and related members of the same family ( Microbacteriaceae ) and class ( Actinobacteria ) is shown in Fig. 2. The general features of the strain are summarized in Table 1. In order to investigate the potential of A. aureus AR33T as plant-associated microbe from a heavy metal contaminated environment, we performed the following additional assays: production of auxins and siderophores, phosphate solubilization, resistance to heavy metals and heavy metal mobilization. To maximize metabolite production necessary for these properties, assays were performed in Landy medium (20 g l−1 glucose, 5 g l−1 glutamate, 0.25 g l−1 MgSO4, 0.25 g l−1 KCl, 0.5 g l−1 KH2PO4, 150 μg l−1 FeSO4, 5 mg l−1 MnSO4, 160 μg l−1 CuSO4, 1 g l−1 yeast extract, pH 7.2) [4], often used for secondary metabolite analysis in gram positive bacteria [5]. The optimal growth temperature and pH values are 28 °C and 6.5–7.5, respectively. AR33T showed oxidase, catalase activity and produced auxins [4, 6]. No phosphate solubilization activity [7] was detected. AR33T is resistant up to 6 mM of zinc and lead and up to 1 mM of cadmium. The production of siderophores was observed using the chrome azurol S assay [8] with Landy (without iron) as growth medium, but not in MM9. The latter is in accordance with a previous study using MM9 [2]. The ability to change the solubility of metals in soil was tested in heavy metal mobilization assays performed as described in [2], but using Landy as growth medium. In Landy, AR33T increased manifold the extractability of lead and iron, whereas the extractability of zinc was slightly increased and the extractability of cadmium, copper and manganese slightly decreased (Fig. 3). Earlier results showed an increase in extractability of both Zn and Cd eased with AR33T in tryptic soy broth [2], suggesting that the production of secondary metabolites such as siderophores and other chelating compounds can be influenced by the growth medium, previously documented for a number of members of the class Actinobacteria [9].

Fig. 1
figure 1

a: Picture of A. aureus AR33T grown in solid Landy medium; b: Confocal laser scanning microscope microphotograph of A. aureus AR33T. Cells were stained with 3 μM green fluorescent nucleic acid stain SYTO9 (ThermoFisher)

Full size image
Fig. 2
figure 2

NJ phylogenetic tree based on 16S rRNA gene sequences. GenBank accession numbers are shown in parenthesis. The sequences were aligned with MUSCLE and the phylogenetic tree was calculated in MEGA6 [34] with bootstrap value of 1000 replicates. In bold red A. aureus AR33T; in red Agromyces spp. with published genomes used in this study for further comparison; in blue and green related members of the Microbacteriaceae family and of the Actinobacteria class used in this study for further comparison, respectively

Full size image
Table 1 Classification and general features of Agromyces aureus AR33T
Full size table
Fig. 3
figure 3

Heavy metal mobilization assays. Contaminated soil was shaken with filtrates of stationary cultures of A. aureus AR33T grown in Landy medium (AR33T, n = 6) and with not inoculated Landy medium (NC, n = 3). Significant differences of culture filtrates to control (p < 0.05 identified with t-tests) are labeled with an asterisk (*). Error bars show the standard error

Full size image

Chemotaxonomic data

A. aureus AR33T has a peptidoglycan type B2γ (D-Glu-L-Dab). Galactose, rhamnose, ribose and fucose constitute the cell-wall sugars. The major cellular fatty acids are anteiso-C15:0, anteiso-C17:0 and iso-C16:0, while diphosphatidylglycerol, glycolipid and phosphatidylglycerol are the predominant polar lipids. The main menaquinones are MK-11, −10 and −12.

Genome sequencing information

Genome project history

The genome of A. aureus AR33T was sequenced by GATC Biotech AG, Konstanz, Germany and subsequently assembled at our institute. The complete genome sequence is available in the NCBI database under the following accession number CP013979. The genome sequencing project information is summarized in Table 2.

Table 2 Project information
Full size table

Growth conditions and genomic DNA preparation

A. aureus AR33T cells were grown in Landy medium for 48 h at 28 °C with continuous shaking at 200 rpm. DNA was isolated using a phenol-chloroform based protocol. Briefly, cells were collected by centrifugation, re-suspended in lysis buffer (0.1 M NaCl, 0.05 M EDTA pH8, lysozyme 100 mg mL−1) and incubated for 10 min at 37 °C. Subsequently, 5% sarkosyl (sodium lauroyl sarcosinate) was added to the solution that was further incubated on ice for 5 min. DNA was extracted using 1 volume of phenol-chloroform-isoamylalcohol (25:24:1) and treated with RNaseA (20 mg mL−1) to remove RNA. After an additional cleaning step with chloroform, the DNA was precipitated using 2.5 volumes of ice-cold absolute ethanol and 0.1 volumes of 3 M sodium acetate (pH 5.2) and incubated for 3 h at −20 °C. Genomic DNA was collected by centrifugation; the pellet was washed with 70% ethanol and re-suspended in water. The quality and quantity of DNA were assessed on 1% agarose gel and measured with the NanoDrop spectrophotometer.

Genome sequencing and assembly

The whole genome was sequenced using the Illumina MiSeq platform (300 bp paired-end reads). Raw reads were screened for PhiX contamination using Bowtie2 [10]. Adapter- and quality-trimming was performed in Trimmomatic-0.32 [11]. Overlapping reads were subsequently merged using FLASH [12] and long single reads and paired end reads assembled with SPAdes 3.1.0 [13]. The initial assembly consisted in 4 contigs, of which one represented the rRNA genes. The gaps between the contigs were closed by designing primers at each contig edge (Additional file 1: Table S1). The PCR products were cloned and sequenced (Sanger). The 4 contigs and the Sanger sequences were manually assembled resulting in a single contig that could be circularized with Circlator [14]. The assembly quality was estimated in QUAST 2.3 [15] and quality control of mapping data performed in Qualimap 1.0 [16]. Phylosift v1.0.1 [17] was used to identify 38 highly conserved, single-copy marker genes that can be used to assess the completeness of the genome [18, 19]. In A. aureus AR33T all marker genes could be identified and the phylogenetic analysis showed no contamination. The presence of tRNA genes for all essential amino acids was verified using ARAGORN [20].

Genome annotation

The A. aureus AR33T genome was annotated using the NCBI Prokaryotic Genome Annotation Pipeline as well as Prokka [21, 22]. BLASTClust [23] was used to detect genes in internal clusters with the following threshold parameters: 70% covered length and 30% sequence identity. The COG functional categories were assigned through the WebMGA server [24]. The predicted CDSs were used to search against the Pfam database [25] to assign them to the corresponding protein families. SignalP [26] and TMHMM [27] were used to identify genes containing signal peptides and transmembrane helices, respectively. The detailed information about these features is summarized in Tables 3 and 4.

Table 3 Genome statistics
Full size table
Table 4 Number of genes associated with general COG functional categories
Full size table

Genome properties

The complete genome of A. aureus AR33T has a total length of 4373124 bp, a CG content of 70.1% and contains three copies of the rRNA operon, of which one has a different 16S rRNA gene sequence (KU141338, KU141339). It has a total of 4005 predicted genes of which 3928 (98.1%) are protein coding genes and 31 are pseudogenes (0.8%). Two thousand nine hundred seventy-nine genes (74.4%) have a functional prediction and 2771 genes (70.5%) could be assigned to a COG functional category (Table 4). Additional information about the genome statistics is shown in Table 3. The map of the genome is represented in Fig. 4.

Fig. 4
figure 4

Graphical circular map of A. aureus AR33T obtained in DNAPlotter [35]. From inner to outer ring: ring 1 GC skew, ring 2 GC% content, ring 3 tRNAs (green), ring 4 rRNAs (blue), ring 5 CDSs on reverse strand (orange) and ring 6 CDSs on forward strand (red)

Full size image

Insights from the genome sequence

To gain more information about the genome of A. aureus AR33T and about the Agromyces genus in general, we performed comparative genomic analysis using other 6 available Agromyces genomes with high quality assembly (Table 5). All genomes were annotated in Prokka [22] and the predicted genes were used in Roary [28] to calculate the Agromyces pan-genome and core-genome. Since these organisms are members of the same genus but belong to different species, we decided to set the Roary minimum blastp percentage identity at 80%. The choice of this threshold value is supported by the bidirectional best hit analysis performed in RAST [29] (Additional file 1: Figure S1). The Agromyces pan-genome has a total of 14,320 genes: 979 represent the core-genome; 3733 and 9608 form the shell and cloud genome, respectively (Fig. 5a). In particular, 1916 genes of A. aureus AR33T have orthologues in the shell genome and 1014 genes seem to be unique (Fig. 5b). Subsequently, we focused our comparative analysis on the two closest related organisms with a publicly available genome: Agromyces sp. Leaf222 and A. italicus DSM 16388 (Fig. 2, Additional file 1: Figure S2). The genome of A. aureus AR33T and Agromyces sp. Leaf222 seem to be the most similar ones having almost half (1575) of their CDSs sharing at least 80% amino acid similarity. Moreover, these two organisms share 137 COG functional categories and 117 KEGG metabolic pathways (Fig. 5c). Despite being part of the same phylogenetic clade (Fig. 2), A. italicus DSM 16388 seems to have a different set of genes and functionalities compared to A. aureus AR33T and Agromyces sp. Leaf222 (Fig. 5c). Finally, a distinctive feature of the A. aureus AR33T genome is the presence of several genes related to metal resistance and homeostasis. For instance, whereas all three have transporters for iron, an essential element, only strain AR33T has transporters also for nickel and cobalt. This feature is probably due to its isolation source, a former zinc/lead mining and processing site, and is in agreement with the displayed ability to mobilize metals (Fig. 3) and to survive in the presence of zinc, lead and cadmium.

Table 5 General features of the genomes of Agromyces spp. and related organisms used for comparative studies
Full size table
Fig. 5
figure 5

a-b Pan-genome of Agromyces spp. calculated in Roary (blastp 80%) [28]. The inner ring shows the total number of the core genes (present in all the species); the middle ring shows the number of genes in the shell of the pan-genome (present in more than one species); the outer rings show the number of genes in the cloud of the pan-genome (present in only 1 species). c Comparison of A. aureus AR33T with the closely related species Agromyces sp. Leaf222 and A. italicus DSM 16388. Venn diagram showing the shared CDSs (Roary, blastp 80%), genes in classified in the same COG functional categories and KEGG metabolic pathways were designed using http://bioinformatics.psb.ugent.be/webtools/Venn/

Full size image

Extended insights

To obtain further insights into the A. aureus AR33T genome, we included related organisms in our comparative analysis: Microbacterium testaceum StLB037, Microbacterium sp. CGR1, Clavibacter michiganensis subsp. michiganensis NCPPB 382, Leifsonia xyli subsp. xyli CTCB07, Cellulomonas flavigena DSM 20109 and Streptomyces coelicolor A3(2). The selection criteria were the following: (i) they have a closed genome; (ii) they are member of the same family ( Microbacteriaceae ) or class ( Actinobacteria ) that have a similar secondary metabolite gene clusters; (iii) they were isolated from soil or are plant-associated bacteria. We performed an all versus all genome comparison in Gegenees [30] to establish the overall similarity of the considered genomes (Additional file 1: Figure S3). The heat map reflects the phylogenetic tree (Fig. 2) and confirms that the closest sequenced relative of A. aureus AR33T is Agromyces sp. Leaf222. Differences between the analyzed genomes are highlighted in the circular map designed in BRIG [31] (Fig. 6). Interestingly, the gaps indicating regions with low similarity to compared genomes correspond to drastic changes in the GC content of A. aureus AR33T and code for: siderophores transporters and biosynthetic clusters, genes related to metal resistance and homeostasis, phage sequences and several hypothetical proteins. A distinctive characteristic of the Actinobacteria class is the ability to produce a wide range of secondary metabolites. Therefore, we identified secondary metabolites gene clusters using antiSMASH 3.0 [32] (Table 6). In the A. aureus AR33T genome, we could identify a type III PKS gene and clusters for the production of terpenoids, siderophores and lantipeptides. The presence of a siderophore biosynthetic cluster is supported by the positive result in the in vitro CAS assay [8] and could explain the ability to change the mobility of metals like iron and lead demonstrated in the heavy metal mobilization assay. This cluster seems to be involved in the production of a desferrioxamine-like siderophore and is found in other members of the Microbacteriaceae family as well. For instance, the genes belonging to the siderophore cluster in Agromyces sp. Leaf222 share 79–98% amino acid similarity with the ones of AR33T. The terpenoid cluster seems to be widespread among these organisms and is often associated to a yellow pigmentation of the colonies. The type III PKS gene shows similarities to a naringenin-chalcone synthase and is conserved among other Agromyces spp. and Microbacteriaceae spp. with the exception of Leifsonia xyli subsp. xyli CTCB07, which has a longer sequence. Finally, the lantipeptide gene cluster is a rare feature and its structure resembles the one that has been characterized in Streptomyces venezuelae for the production of lanthionine-containing peptides [33].

Fig. 6
figure 6

Circular visualization of the whole genome comparison of A. aureus AR33T, other Agromyces spp., related members of the same family and class. The figure was designed using BRIG [31]. The gaps in the circles represent regions of low or no similarity and contain the following features: (1) siderophore biosynthetic gene cluster (desferrioxamine-like); (2) metal related genes like transporters for Pb/Cd/Zn/Hg and for the resistance to As; (3) non-ribosomal peptide synthase modules (pyoverdine-like siderophore) and ABC siderophore transporters; (4) several hypothetical proteins, Mg/Co/Ni transporters, Co/Zn/Cd resistance genes; (5) several hypothetical protein and phage sequences that were detected also in PHAST [36]; (6) Co/Ni transporters, pathway for aromatic compound degradation, transporters for branched chain amino acids; (7) Pb/Cd/Zn/Hg transporters, resistance genes for Cu/Co/Zn/As/Cd, a phage integrase; (8) genes for the production of exopolysaccharides; (9) several hypothetical proteins; (10) Na + H+ antiporters; (11) ABC transporters for Co and heme/siderophore complexes

Full size image
Table 6 Secondary metabolite gene clusters identified with antiSMASH [32] in the genomes Agromyces spp. and related organisms. Others: cluster containing a secondary metabolite-related protein that does not fit into any other antiSMASH category; putative: putative cluster identified with the ClusterFinder algorithm which is mainly related to saccharides or fatty acids or without a specific prediction
Full size table

Conclusions

Heavy metals are recognized as one of the main soil contaminants world-wide. Bacteria such as A. aureus AR33T could be used to improve eco-friendly decontamination techniques such as bio-augmentation or phytoremediation. Here, we presented the first complete genome of an Agromyces that was isolated from a heavy metal mining/processing site in Austria. It is able to survive in the presence of metals such as zinc, lead and cadmium and can influence the metals mobility of a contaminated soil. Genomic analysis revealed the presence of secondary metabolite gene clusters potentially involved in terpenoid and lantipeptide production, type III PKS and siderophore biosynthesis. In particular, the last two gene clusters could be directly involved in the heavy metal im-mobilization process. Moreover, the correlation between the genotype and phenotype of A. aureus AR33T is supported by the presence of several metal resistance and homeostasis genes. We could identify genomic regions displaying low similarity to compared genomes of related organisms, which are characterized by a different GC content and by the presence of genes coding for siderophore transporters and biosynthetic clusters, genes related to metal resistance and homeostasis, phage sequences and several hypothetical proteins. The genome based phylogenetic analysis including closely related and more distant organisms isolated from similar environments appeared to be in agreement with the 16S rRNA gene phylogeny. This brief comparative analysis could be the starting point for further studies in different directions. For instance, it could lead to a deeper understanding of the Agromyces genus and its relationship with other members of the class Actinobacteria and to a better knowledge about the im-mobilization mechanisms.